Skip to main content
SpringerLink
Log in

A Temporal Topic Model for Noisy Mediums

  • Conference paper
  • First Online:
Advances in Knowledge Discovery and Data Mining (PAKDD 2018)

Part of the book series: Lecture Notes in Computer Science ((LNAI,volume 10938))

Included in the following conference series:

Abstract

Social media and online news content are increasing rapidly. The goal of this work is to identify the topics associated with this content and understand the changing dynamics of these topics over time. We propose Topic Flow Model (TFM), a graph theoretic temporal topic model that identifies topics as they emerge, and tracks them through time as they persist, diminish, and re-emerge. TFM identifies topic words by capturing the changing relationship strength of words over time, and offers solutions for dealing with flood words, i.e., domain specific words that pollute topics. An extensive empirical analysis of TFM on Twitter data, newspaper articles, and synthetic data shows that the topic accuracy and SNR of meaningful topic words are better than the existing state.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution
Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 84.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 109.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Notes

  1. 1.

    Another way to simulate this is to sample from a Zipfian distribution. Our data generator allows for distribution changes. For these experiments, we create a mixture that is noisier and harder to generate topics from than a Zipfian sample.

References

  1. de Arruda, H.F., da Fontoura Costa, L., Amancio, D.R.: Topic segmentation via community detection in complex networks. CoRR (2015). http://arxiv.org/abs/1512.01384

  2. Bhadury, A., Chen, J., Zhu, J., Liu, S.: Scaling up dynamic topic models. In: WWW. SIAM (2016)

    Google Scholar 

  3. Blei, D.M., Lafferty, J.D.: Dynamic topic models. In: ICML. IEEE (2006)

    Google Scholar 

  4. Blei, D.M., Lafferty, J.D.: Visualizing topics with multi-word expressions. arXiv e-prints (2009)

    Google Scholar 

  5. Blei, D.M., Ng, A.Y., Jordan, M.I.: Latent Dirichlet allocation. J. Mach. Learn. Res. 3, 993–1022 (2003)

    MATH  Google Scholar 

  6. Blondel, V.D., Guillaume, J.-L., Lambiotte, R., Lefebvre, E.: Fast unfolding of communities in large networks. J. Stat. Mech. Theor. Exp. 2008(10), P10008 (2008)

    Article  Google Scholar 

  7. Cataldi, M., Di Caro, L., Schifanella, C.: Emerging topic detection on Twitter based on temporal and social terms evaluation. In: MDM-KDD. ACM (2010)

    Google Scholar 

  8. Chang, J., Boyd-Graber, J., Wang, C., Gerrish, S., Blei, D.M.: Reading tea leaves: how humans interpret topic models. In: NIPS. AAAI (2009)

    Google Scholar 

  9. InternetLiveStats: Twitter usage statistics. http://www.internetlivestats.com/twitter-statistics/. Accessed 05 May 2017

  10. Kasiviswanathan, S.P., Melville, P., Banerjee, A., Sindhwani, V.: Emerging topic detection using dictionary learning. In: CIKM. ACM (2011)

    Google Scholar 

  11. Lafferty, J.D., Blei, D.M.: Correlated topic models. In: NIPS, pp. 147–154. AAAI (2006)

    Google Scholar 

  12. Noyes, D.: The top 20 valuable Facebook statistics - updated May 2017. https://zephoria.com/top-15-valuable-facebook-statistics/. Accessed 05 May 2017

  13. Shahnaz, F., Berry, M.W., Pauca, V., Plemmons, R.J.: Document clustering using nonnegative matrix factorization. Inf. Process. Manage. 42, 373–386 (2006)

    Article  Google Scholar 

  14. Sleeman, J., Halem, M., Finin, T., Cane, M., et al.: Modeling the evolution of climate change assessment research using dynamic topic models and cross-domain divergence maps. In: Symposium on AI for Social Good. AAAI (2016)

    Google Scholar 

  15. Teh, Y.W., Jordan, M.I., Beal, M.J., Blei, D.M.: Hierarchical Dirichlet processes. J. Am. Stat. Assoc. 101(476), 1566–1581 (2006)

    Article  MathSciNet  Google Scholar 

  16. Varol, O., Ferrara, E., Davis, C.A., Menczer, F., Flammini, A.: Online human-bot interactions: detection, estimation, and characterization. In: ICWSM. AAAI (2017)

    Google Scholar 

  17. Wang, X., McCallum, A.: Topics over time: a non-Markov continuous-time model of topical trends. In: KDD. ACM (2006)

    Google Scholar 

  18. Yan, X., Guo, J., Liu, S., Cheng, X., Wang, Y.: Learning topics in short texts by non-negative matrix factorization on term correlation matrix. In: SDM. SIAM (2013)

    Chapter  Google Scholar 

Download references

Acknowledgements

This work was supported by the Massive Data Institute (MDI) at Georgetown University.

Author information

Authors and Affiliations

  1. Georgetown University, Washington, D.C., USA

    Rob Churchill, Lisa Singh & Christo Kirov

Authors
  1. Rob Churchill
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Lisa Singh
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Christo Kirov
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Rob Churchill .

Editor information

Editors and Affiliations

  1. Deakin University, Geelong, Victoria, Australia

    Dinh Phung

  2. National Chiao Tung University, Hsinchu City, Taiwan

    Vincent S. Tseng

  3. Monash University, Clayton, Victoria, Australia

    Geoffrey I. Webb

  4. Japan Advanced Institute of Science and Technology, Nomi, Ishikawa, Japan

    Bao Ho

  5. University of Melbourne, Melbourne, Victoria, Australia

    Mohadeseh Ganji

  6. University of Melbourne, Melbourne, Victoria, Australia

    Lida Rashidi

Rights and permissions

Reprints and permissions

Copyright information

© 2018 Springer International Publishing AG, part of Springer Nature

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Churchill, R., Singh, L., Kirov, C. (2018). A Temporal Topic Model for Noisy Mediums. In: Phung, D., Tseng, V., Webb, G., Ho, B., Ganji, M., Rashidi, L. (eds) Advances in Knowledge Discovery and Data Mining. PAKDD 2018. Lecture Notes in Computer Science(), vol 10938. Springer, Cham. https://doi.org/10.1007/978-3-319-93037-4_4

Download citation

  • DOI: https://doi.org/10.1007/978-3-319-93037-4_4

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-319-93036-7

  • Online ISBN: 978-3-319-93037-4

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics

Search

Navigation